Last week’s discussion of the worst pitchers started by defending World Series champions on Opening Day inspired me to look into the worst Opening Day starters period, regardless of where his team had finished the previous season. Claude Osteen, whom the Dodgers trotted out in Game One of their World Series title defense on April 12, 1966, doesn’t even sniff the title of “worst Opening Day starter ever”:
Worst Opening Day Starters Since 1951
Year |
Team |
Previous Year WARP |
|
1969 |
California Angels |
-2.128 |
|
1981 |
-1.481 |
||
1975 |
-1.381 |
||
1977 |
Randy Jones |
San Diego Padres |
-1.264 |
1970 |
Montreal Expos |
-1.201 |
|
1982 |
Randy Jones |
-1.180 |
|
1980 |
Randy Jones |
San Diego Padres |
-1.082 |
1983 |
-1.053 |
||
1964 |
-0.999 |
||
2010 |
-0.998 |
Osteen’s 0.027 WARP in 1965 ranks as the 94th-worst previous-year WARP for an Opening Day starter since 1951, and he is one of 11 pitchers to appear multiple times among the dregs of the bottom 100.
Pitchers Making Multiple Appearances Among Worst 100 Opening Day Starters Since 1951
Appearances |
|
Randy Jones |
5 – 1975 Padres, 1976 Padres, 1977 Padres, 1980 Padres, 1982 Mets |
4 – 1973 Expos, 1976 Athletics, 1977 Athletics, 1978 Red Sox, 1984 Mets |
|
2 – 1975 Tigers, 1976 Tigers |
|
2 – 1954 Pirates, 1956 Pirates |
|
Rick Langford |
2 – 1982 Athletics, 1983 Athletics |
2 – 2009 Nationals, 2010 Nationals |
|
Ken McBride |
2 – 1963 Angels, 1964 Angels |
2 – 1994 Tigers, 1995 Tigers |
|
Claude Osteen |
2 – 1966 Dodgers, 1964 Senators |
2 – 1952 Cardinals, 1953 Cardinals |
|
2 – 1967 Yankees, 1969 Yankees |
With that out of the way, let's move on to new business. Today we'll take a look at pitchers piling up wins but not throwing enough innings to qualify for the ERA title, organization depth as a precursor to spring training success, and the worst Opening Day lineups of the modern era. If you have a question you would like to have answered, please send me an e-mail or reach out via one of the myriad contact links on this page (be sure to include your full name and hometown).
How many pitchers who didn't throw enough innings to qualify for the ERA title have placed among the top five in wins in a given MLB season?
Kyle Eliason
St. Paul, MN
Since 1950, it has taken an average of 17.387 wins to crack the top five, a feat that has grown more difficult as the role of the bullpen has evolved, resulting in shorter outings by starters and fewer innings per appearance by relievers. Given that, it’s no wonder 30 years have passed since Bob Stanley’s 10 wins out of the Red Sox bullpen in 1981 earned him a tie with nine others for the fifth-most wins in baseball.
Pittsburgh reliever Roy Face compiled an 18-1 record in 57 appearances in 1959, tying with Vern Law and Bob Shaw for fourth place. Face was one of seven players to pitch in 50 or more games without making a single start that year, and one of only three to do it in both 1958 and 1959 (Don Elston and the aforementioned Gerry Staley were the others).
This year's top 4 spring training teams (St. Louis, San Diego, Oakland, and Toronto) all have strong farm systems. Is it a case of "our AA and AAA guys are crushing yours" or is it just "we attract more interesting former major leaguers than you do" or is it completely random? Is the top four's correlation with farm talent just a coincidence this year?
John Franco
Pittsburgh, PA
Each March, we’re reminded that spring training player statistics and team won-lost records are poor predictors of performance in the upcoming season. For the most part, that advice is correct.

Between 2006-2011, the difference between a team’s spring training and regular season winning percentages has ranged wildly, from -0.256 to 0.255, reaffirming conventional wisdom that spring training won-lost records are meaningless.
John’s question of whether there is correlation between farm system strength and team spring training performance is something I hadn’t considered before. To test his theory, I turned to the organization talent rankings Kevin Goldstein has compiled each offseason since 2007. I compared Kevin’s ranking of an organization with its spring training winning percentage, but found little evidence that system-wide talent depth played a part in a team’s spring training success.

The trend is more apparent when the data is aggregated by rank, but still isn’t enough to support any conclusions. I suspect stronger correlation could be found if rankings of upper-level, rather than organizational, talent were used, however.
In 2012, Kevin had San Diego and Toronto atop his rankings, but for different reasons. Five of the Padres’ top 11 prospects played at or above Double-A last year, while nine of the Blue Jays’ top 11 saw time at Class-A or below. The Padres boast a deep supply of upper-level talent that is likely to reach the majors, though unlikely to develop into stars. Contrast that with Toronto, whose farm is shallower (but not shallow) at the top but includes several prospects with star potential at various levels within the organization.

As it is with September stats, numbers accrued during spring training are subject to far too many variables to be useful on either end of a cause-effect inquiry. Aside from the ranging levels of talent deployed, pitchers are not at full strength when spring games begin and only throw full outings in the days leading up to the start of the regular season. Because spring training games don’t count, a pitcher is free to experiment and tinker with grips, mechanics, and new pitches altogether, further diminishing the reliability of the data earned by players and teams.
Which team has had the most potential in history? That is, the one with the most future WARP from its opening day roster? What team had the least?
David Kociemba
Watertown, MA
Unfortunately, I don’t have each team’s Opening Day roster available. What we can do, however, is look at which teams have had the most (and least) future WARP in their Opening Day lineups. We’ll cut the search off at the year 2000 so as to have a healthy number of seasons for each lineup to work with.
The starting nine fielded by the Milwaukee Braves on Opening Day 1955 featured six players who would go on to be worth at least 15 wins above replacement from that day forward, including a 21-year-old Henry Louis Aaron:
Milwaukee Braves Opening Day Lineup, April 12, 1955
Lineup |
Player |
Future WARP |
1 |
31.05 |
|
2 |
147.88 |
|
3 |
73.94 |
|
4 |
8.24 |
|
5 |
22.62 |
|
6 |
17.36 |
|
7 |
-0.72 |
|
8 |
22.99 |
|
9 |
12.39 |
Aaron and Mathews carried the Braves to the top four spots on our list, and the 1960-61 San Francisco Giants—led by Willie Mays, Willie McCovey, and Orlando Cepeda—are the only other franchise to appear multiple times among the top eight.
Best Opening Day Lineups by Future WARP
Rank |
Year |
Team |
Future WARP |
1 |
1955 |
Milwaukee Braves |
335.73 |
2 |
1954 |
Milwaukee Braves |
330.08 |
3 |
1956 |
Milwaukee Braves |
320.51 |
4 |
1957 |
Milwaukee Braves |
292.99 |
5 |
1961 |
San Francisco Giants |
290.77 |
6 |
1963 |
277.14 |
|
7 |
1991 |
267.43 |
|
8 |
1960 |
San Francisco Giants |
263.21 |
9 |
1976 |
260.27 |
|
10 |
1992 |
Houston Astros |
260.03 |
The Seattle Mariners rolled out the worst Opening Day lineup of all time—at least in terms of future WARP—on April 9, 1980. Only first baseman Bruce Bochte (7.14) and second baseman Julio Cruz (5.55) performed at an above-replacement level for the remainder of their careers; the rest of the Mariners lineup combined for -6.46 WARP.
Opening Day in the Devil Rays' inaugural season is worth a mention, but not because of how bad Tampa Bay’s lineup was; rather, how spectacular it used to be. Hometown heroes Wade Boggs and Fred McGriff combined for 112.2 career WARP prior to teaming up in the heart of the Devil Rays’ lineup in 1998. The two would be worth only 5.8 WARP over the next nine seasons.
Worst Opening Day Lineups by Future WARP
Rank |
Year |
Team |
Future WARP |
1156 |
1980 |
Seattle Mariners |
6.23 |
1155 |
1998 |
Tampa Bay Devil Rays |
7.57 |
1154 |
1988 |
8.17 |
|
1153 |
1982 |
Seattle Mariners |
10.28 |
1152 |
2000 |
Tampa Bay Devil Rays |
13.11 |
1151 |
1963 |
New York Mets |
14.88 |
1150 |
1961 |
Washington Senators |
16.48 |
1149 |
1980 |
New York Mets |
16.54 |
1148 |
1974 |
16.59 |
|
1147 |
1977 |
16.81 |
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.
Subscribe now
I know that each system computes their numbers differently so there will be differences but this seems kind of extreme.
Colin, or whoever at BP: we really, really, really need a detailed assessment of FRA, and comparisons with FIP, SIERA, etc. It could be right, but it spits out enough head-scratching valuations that I can't really trust it without more convincing (See also: Cain, Matt and Hernandez, Felix).
1. Chris Short (1970, PHI) - 7.77 FRA
2. Darryl Kile (2000, SLN) - 7.29
3. Mike Moore (1995, DET) - 7.13
4. Randy Jones (1982, NYN) - 6.94
5. Bud Podbielan (1953, CIN) - 6.81
6. Kevin Ritz (1997, COL) - 6.70
7. Mario Soto (1988, CIN) - 6.60
8. Fausto Carmona (2010, CLE) - 6.57
9. Scott Schoeneweis (2001, ANA) - 6.52
10. Terry Mulholland (1995, SFN) - 6.49
I understand that the BBWA gets awards wrong, but I'm pretty sure they've never given a Cy Young award to a below replacement level pitcher.
Jones inspired a bit of discussion at BP today, and I'll paste some of what Rob McQuown had to say here:
"I think that while we may indeed find flaws in the system, in Jones case, there will always be things are are difficult for people to wrap their intuitive minds around, given the 'Old School' mindset of ERA (or wins or whatever) being a prime indicator of pitcher goodness. In Jones' case, he combined fairly high levels of unearned runs (i.e. fielders making errors behind him) with BABIPs that were good (sometimes exceptional), compared to the league BABIPs. So, ERA double-credits him for his fielders, both removing any blame for errors behind him (unearned runs), and ALSO giving him full credit for the range of the players (lower BABIP than average fielding would provide).
FAIR_RA (or FRA) theoretically removes all of this from the pitcher's hands - instead utilizing league averages for turning these balls into outs. It's more complicated than that, but I do think that his low VORPs and WARPs (driven by high FRAs) are demonstrative of something that can be explained and could prove educational to thinking baseball fans.
Whatever I find in the WARP formulae, probably the big bone of contention will be the starting assumption that only "three true outcomes" are under a pitcher's control (essentially), and the rest is just environment/fielding/luck."
It seems like FRA is punishing him (brutally) for not being a high strikeout pitcher. He was a ground ball pitcher who didn't walk guys and kept the ball in the park. I'm not sure FRA is the best measure here.
Put it this way, if you were the General Manager of the Padres in the 70's, would you waive this guy? He's below replacement level...
It's not like that was some one-year BABIP fluke or anything; dude's run prevention was pretty solid throughout his career.
At any rate, FRA does a lot more than just look at TTO, right? Looking at 1976 vs. 1978, his FIPs are quite similar, why is FRA a full run lower in 78? I notice his BABIP was quite a bit higher in 78 than 76, so given similar FIP* it might attribute more run prevention that year to the pitcher....but does that really make sense? If the things with known good year-to-year and run prevention correlations are similar, why would *increasing* the rate of hits allowed *reduce* the number of runs attributed to him?
*the other thing of course is that his HR/9 is way lower in 1978, and I wrote a long thing about that before noticing the BABIPs, but long story short K and UBB rates were better in 76 so FIP works out to be similar, and I estimate it should only change his RA by at most about .3, not 1.0, so unless his HR rate is in a denominator somewhere that shouldn't be it...
There's no way any starting pitcher throws 315 innings with a WHIP of 1, with 5 shutouts and an ERA more than a run below the league average, and measures as below replacement level.
With certain anomalous guys, maybe they provide an opportunity to figure out how the stat can be better calibrated, rather than assume that what seems to be true to everyone else---I mean, actual results do mean something, especially over a large sample---is in fact wrong, at least according to the formula.
If the Padres' defense of that era was so superior that it made a sub-replacement pitcher look like an ace, and none of that out creation was really to Jones's own credit, then the Padres' staff should have had two or three apparent aces every year, at least. This was not the case.